viewpoint
viewpoint
Are microbes at the root of a solution
to world food production?
Rational exploitation of interactions between microbes and plants can help to transform agriculture
John P. Morrissey, J. Maxwell Dow, G. Louise Mark & Fergal O’Gara
O
ne major challenge for the twenty-
first century will be the produc-
tion of sufficient food—the
United Nations Population Fund estimates
that the global human population may
well reach 10 billion by 2050
(www.unfpa.org). This means increasing
agricultural productivity of food crops, as
plants form the basis of every food chain.
However, agriculture in developed coun-
tries already creates a range of serious
environmental problems through the use
of pesticides and herbicides, salinization
and the depletion of water resources.
Furthermore, agricultural production in
many African, Asian and South American
countries cannot be increased without fur-
ther destroying more areas by turning
them into arable land, thus threatening
global biodiversity, which is already under
stress from human action. If global food
production is to keep pace with an
increasingly urbanized and growing
population, while formulating new food
production strategies for developing coun-
tries, the great challenge for modern soci-
eties is to boost plant productivity in an
environmentally sustainable manner.
A major effort in plant biotechnology is
therefore the development of new crop
varieties with enhanced disease and pest
resistance, greater drought and salt toler-
ance and better nutritional value through
the introduction of desirable traits either
by conventional breeding or genetic modi-
fication. However, these efforts have tradi-
tionally focused on plant phenotypes.
What has been largely ignored is the
important role of microbial communities
9 2 2 EMBO reports VOL 5 | NO 10 | 2004
that interact with plants to influence plant
health, productivity and biodiversity. The
impact of the microbial world on plants is
evident: worldwide each year, microbial
diseases cost crop producers billions of
Euros. Similarly, the important role of
nitrogen fixation by rhizobia and other
bacteria for plant growth has been known
for decades. What is less appreciated, and
less well understood, is the pervasive
influence that other microbes have on
plant health and growth; they enhance
stress tolerance, provide disease resis-
tance, aid nutrient availability and uptake,
and promote biodiversity. Although major
advances in genomic technologies and
in situ studies of beneficial plant–microbe
interactions have produced a large
amount of knowledge and given insights
into the mechanisms of these interactions,
their application in biotechnology and
agriculture has yet to be exploited. A
greater understanding of how plants and
soil microbes live together and benefit
each other can therefore provide new
strategies to improve plant productivity,
while helping to protect the environment
and maintain global biodiversity.
T
he most intense interactions
between microbes and plants take
place at the rhizosphere, which is
the interface between plant roots and the
soil. The concept of the rhizosphere was
broached 100 years ago by the German
biologist Lorenz Hiltner (Fig 1), who first
proposed that the area around the roots is
a region of high microbial activity
(Hiltner, 1904). Despite a large amount of
©2004 EUROPEAN MOLECULAR BIOLOGY ORGANIZATION
If global food production is to
keep pace with an increasingly
urbanized and growing
population […] the great
challenge for modern societies
is to boost plant productivity
in an environmentally
sustainable manner
work and research on the biology and
biochemistry of the rhizosphere, there is
still very little known about the detailed
mechanisms that take place in the soil.
Although most plant biodiversity research
focuses on what is visible above the
ground, researchers have demonstrated a
clear link between ‘above-ground’ and
‘below-ground’ diversity (Helgason et al,
1998; Marcel et al, 1998). The biological
and ecological details that underpin this
phenomenon remain elusive, but these
findings suggest that maintenance of a
high diversity of plant species requires
a correspondingly high level of diversity
in the soil microbial community (Wardle
et al, 2004).
The influence of plants on microbial
population structure and function in the
rhizosphere has important ecological
implications for soil function, including
biogeochemical cycles. Similarly, soil
microbes have a tremendous influence on
plant health and productivity (Bloemberg
& Lugtenberg, 2001). One straightforward
and visible benefit for the plant is a better
supply of and access to nutrients. The role
of mutualistic nitrogen-fixating rhizobia
v iew point
Fig 1 | Lorenz Hiltner: pioneer of the rhizosphere concept. Courtesy of the
Bavarian State Institute for Agriculture in Freising, Germany
has been well documented for decades,
but recent data detail the intimate
exchange of nutrients during the symbiosis
of plant roots and bacteria (Lodwig et al,
2003). The plant attracts nitrogen-fixating
bacteria to invade the cells in the root and
provides them with carbohydrates as a
food source while the bacteria reduce
nitrous compounds in the soil that are then
used by the plant. Similarly, interactions
science & society
significant environ- agriculture at present. A reduction in the
mental benefits as use of these chemicals would lead to obvi-
they would allow a ous environmental benefits and it would
reduction in the also appeal to consumers who seek more
application of nitro- natural produce.
gen and phospho-
rous fertilizers. The What has been largely ignored is
overuse of such fer-
tilizers has become
the important role of microbial
a major concern communities that interact with
because they cause plants to influence plant health,
nitrate contamina- productivity and biodiversity
tion of soil and
groundwater by
leachates and In addition to these direct effects on
because microbial plant growth, rhizobacteria exert another
denitrification con- health-promoting effect on the plants with
verts residual which they interact. This phenomenon is
nitrogen into the known as induced systemic resistance
greenhouse gas and arises when interactions with non-
nitrous oxide pathogenic bacteria confer better disease
(Nosengo, 2003; resistance on plants (van Loon et al,
Reay, 2004). Equally, 1998). Induced systemic resistance differs
excess phospho- from acquired systemic resistance in
rous compounds plants; the latter occurs in response to
leach into ground- localized microbial attack. Induced resis-
water, rivers and tance is triggered by microbial inter-
streams, where they actions in the rhizosphere, but also
promote algal growth enhances resistance of remote aerial plant
and other environ- parts against pathogens. This systemic
mental problems. change in plant physiology bears con-
Although commer- ceptual similarities to other phenotypes,
cial fertilizer prod- such as improved stress, drought or dis-
ucts based on ease resistance, that have been linked to
rhizobia and plant–fungal associations. Little is known
Azospirillum are already available, their about why bacteria induce this condition,
wider application is restricted by inconsis- which is clearly beneficial to the plant.
tent performance and limitations of host Although induced systemic resistance is
range. A better understanding of plant– generally ascribed to interactions with
microbe symbiosis could help to overcome bacteria, there is emerging evidence that
at least some of these problems. some fungi, particularly endophytic fungi,
In addition to enhancing the nutrient may also be able to induce a similar
supply of plants, microbes also confer a response in plants (Arnold et al, 2003;
degree of protection against plant diseases. Harman et al, 2004; Clay, 2004).
In particular, various bacteria and fungi—

between plants and fungi can also provide
nutrients for the plant. Arbuscular mycor-
rhizal fungi, which form an intricate inter-
nal symbiosis with the roots of most
flowering plants, are associated with the
provision of phosphorous to the plant in
exchange for organic carbohydrates
(Smith & Read, 1997). Microbes also indi-
rectly aid nutrient uptake—bacteria of the
Azospirillum genus promote increased
root mass and more efficient nitrogen
uptake from the soil in response to the
plant hormone indole-3-acetic acid. Using
these bacteria and fungi could provide
especially of the genera Pseudomonas,
Bacillus and Trichoderma—produce a
range of metabolites against other phyto-
pathogenic fungi (Bloemberg
Lugtenberg, 2001; Walsh et al, 2001;
Raijaamkers et al, 2002). Such biocontrol
agents are already efficiently used in the
field, but they have not yet attained the
degree of efficacy and consistency that is
needed for large-scale commercialization.
But there is the potential for improvement
and, with development, such microbes
could become a realistic alternative to
the heavy fungicide regimens used in
T
he interactions of rhizosphere
microbes with plants depend on the
establishment of intimate associa-
& tions between the two partners (Fig 2).
Research on some of these interactions,
such as those between symbiotic rhizobia
and legumes, has demonstrated that this
intimate cooperation between plant and
bacteria displays a high level of host
specificity. There is also a growing body
of evidence suggesting that many other
associations between plants and microbes
show similar degrees of specificity;
different plant species, and even different

©2004 EUROPEAN MOLECULAR BIOLOGY ORGANIZATION EMBO reports VOL 5 | NO 10 | 2004 9 2 3

 science & society
cultivars of the same plant species, estab-
lish distinct microbial populations in their
rhizospheres when grown in the same
soil. The formation of these communities
depends, at least in part, on the activation
of specific programmes of gene expres-
sion in the microbe in response to chemi-
cal signals secreted from the plant. A
pertinent example is the induction of
nodulation genes in receptive rhizobia,
which are triggered by the production and
secretion of particular flavonoids by the
plant. In the case of the rhizobia–legume
interaction, the plant also responds to
bacterial signals, and it is likely that this
type of chemical cross-talk is typical of
other microbe–plant interactions. Other
examples of plant-derived signals that
influence microbial gene expression
include phenolics exuded from plant
wounds, which induce expression of viru-
lence genes in pathogenic Agrobacterium
spp., and compounds that mimic the
quorum sensing signals used by bacteria
to regulate gene expression (Loh et al,
2002; Newton & Fray, 2004). In general,
however, there is only scant knowledge of
signalling interactions between beneficial
microbes and plants. Understanding how
microbes respond to plant signals in terms
of growth and gene expression, and the
role that plant signalling has in determin-
ing interaction specificity or driving popu-
lation selection is central to reaping the
benefits of plant–microbe interactions.
The influence of plants on
microbial population structure
and function in the rhizosphere
has important ecological
implications for soil function,
including biogeochemical cycles
Before the advent of genomic tech-
nologies, scientists had only limited
options to investigate these interactions in
detail, particularly with a view to their
commercial exploitation. This situation
has changed considerably, now that
the genomes of more than ten plant-
associated bacteria have been sequenced,
and the sequencing of another 35 relevant
genomes is already under way (Puhler
et al, 2004). A detailed investigation of the
molecular basis of pathogenic, symbiotic
and associative plant–microbe interac-
tions, both at the levels of comparative
9 2 4 EMBO reports VOL 5 | NO 10 | 2004
Fig 2 | Intimate associations between bacteria and
plants: fluorescently labelled bacteria visualized
in situ growing on the root surface (unpublished
figure provided by Estibaliz Larrainzar)
and of functional genomics is now possi-
ble. So far, most data have come
from comparative analyses with a particu-
lar emphasis on mutualistic and patho-
genic interactions. Although comparative
genomics focuses on genetic potential
rather than gene expression, these studies
do raise some interesting questions about
the distinction between pathogenic and
non-pathogenic bacteria. For instance,
investigations have already yielded a sur-
prising insight regarding type-III protein
secretion systems. These are ordinarily
associated with pathogenicity in bacteria
and are possibly also involved in non-
pathogenic associations (Puhler et al,
2004). In the future, transcriptome profil-
ing and functional genomics (Fig 3) are
likely to produce more information about
microbial responses to plant signals and
the contribution of specific gene products
to the establishment of an interaction with
the host. An important question is
whether plant metabolites exuded from
roots induce microbial type-III protein
secretion systems, and, if so, how?
Pseudomonas fluorescens, for instance, is
chemotactic towards components of root
exudates (de Weert et al, 2002), but the
global effect of these exudates on gene
expression in the bacterium is still
©2004 EUROPEAN MOLECULAR BIOLOGY ORGANIZATION
v iew point
unknown. By profiling the complete
genetic response to root exudates, it will
be possible to assemble a full picture of
how gene expression, and thereby func-
tion, is modulated in the bacterium after
perception of plant signals. It will facili-
tate studies of how different bacterial
species and subspecies respond to partic-
ular plants and how a bacterium responds
differentially to signals from different
plant species or varieties. This, in turn,
will lead to a better understanding of the
basis of host specificity and host selection
during microbe–plant interactions. It may
also be possible to describe general prin-
ciples of plant–microbe communication
that define or distinguish associative,
mutual and pathogenic interactions of
microbes with plants.
In addition to the economic effects of
plant disease, microbial interaction with
plants can also have serious and direct
consequences for human health. One
notorious case is the reputed association
between the consumption of ergot-
contaminated products and the hysteria
that characterized the Salem witch trials in
seventeenth century USA. A hypothesis has
been put forward that psychoactive alka-
loids related to LSD, produced by the
fungus Claviceps purpurea during rye infec-
tion, caused the symptoms seen during that
period, although this can never be un-
equivocally proven. Another example of
microbial toxins in the food chain are fun-
gal aflatoxins, produced by certain strains
of Aspergillus flavus and A. parasiticus
growing on foods and feeds, which are a
serious health concern. The effects of
microbes on plant physiology itself have
received less attention, although it is known
that many food crops naturally produce
toxic metabolites; potatoes and tomatoes,
for example, can accumulate high levels of
toxic steroidal alkaloids (Friedman, 2002;
Korpan et al, 2004). The possible role of
microbes in inducing toxin production, or
in modifying metabolites or metabolic
pathways within the plant, remains largely
unexplored. There are clear precedents for
this premise, for instance the production of
secondary metabolites by plants, such as
phytoalexins, in response to pathogenic
Understanding how microbes
respond to plant signals […] is
central to reaping the benefits of
plant–microbe interactions
v iew point
Fig 3 | The ‘-omics’ technologies: proteomics and genomics are revolutionizing the study of interactions
between plants and microbes (unpublished figure provided by Olive Gleeson and Gordon McAlester)
attack and fungal modification of plant
saponins (Morrissey & Osbourn, 1999;
Bouarab et al, 2002). Although it is clearly a
complex issue, the combination of plant
and microbial functional genomics with
metabolome analysis provides a route to
start addressing these questions.
F
uture biotechnological developments
in the agricultural sector—whether
based on gene modification tech-
nology or on traditional breeding—should
recognize the importance of plant–microbe
associations. The ‘traditional’ plant bio-
technology sector is based around plant
breeding and the selection of varieties with
desired traits, and it pays little attention to
plant–microbial ecology. But the expres-
sion of desirable traits, such as disease
resistance, or drought and salt tolerance,
could equally be driven by interactions
between a particular plant variety and the
colonizing microbial flora. Conversely,
particular plant genotypes may attract a
microbial flora with undesirable traits.
Understanding the genetic basis of
plant–microbe interactions in the context of
how the plant selects its microbial popula-
tion in the soil may allow ‘conditioning’ of
the rhizosphere to promote desirable traits
in the plant, which is, in fact, the basis of
natural disease-suppressive soils. Similarly,
it is also an attractive proposition to
influence or modulate the microbes that
©2004 EUROPEAN MOLECULAR BIOLOGY ORGANIZATION
interact with the plant to generate improved
productivity or health for the latter.
Furthermore, bacteria could be genetically
engineered to confer increased disease
resistance or growth promotion that is only
activated when the bacterium is associated
with its host plant. From the perspective of
developing nations, these are exciting
strategies that may help to increase yields
while avoiding some of the costs and
environmental problems that come with
the use of fertilizers, pesticides, herbicides
and fungicides. However, most of the
microbial biodiversity in soil remains
unexplored and much work remains to be
done to first identify and then characterize
microorganisms that could be used
in such applications. Furthermore, such
approaches require a detailed knowledge
of the molecular signalling that takes place
between plants and microbes to drive
expression of desirable traits and suppress
unwanted effects in a controlled manner.
Exploiting plants and microbes by using
such an integrated approach requires a sys-
tems biology strategy to understand the
degree and complexity of plant–microbe
interactions through the application of
modern ‘-omics’ technologies.
One could describe the evolution of
technology as passing through three phases:
the low-tech stage, the advanced stage and
the high-tech stage. The low-tech phase is,
as the name implies, characterized by the
science & society
Applying knowledge about
beneficial plant–microbe
interactions […] may allow us
to increase food production
while reducing stress on the
environment and on global
biodiversity
use of basic technology; for instance,
before the advent of vaccines and anti-
biotics, care and hygiene were the only
means to help patients to cope with infec-
tious diseases. The advanced phase uses
modern technologies but is clunky and
intrusive, such as the infamous iron lung
that was used in the 1950s and 1960s to
help polio victims breathe. The high-tech
stage finally provides elegant, efficient and
seemingly simple solutions on the basis of
detailed scientific knowledge—by under-
standing the problem and addressing the
cause rather than the symptoms. In the
fight against polio, it was the Salk and
Sabin vaccines that eventually helped the
medical community to combat the disease.
Modern agriculture has gone through
similar phases in recent history. The first
agricultural revolution in the eighteenth
century introduced crop rotation to take
advantage of and manipulate microbial
populations in the soil, although at that
time it was not known why this benefited
plant health and growth. The second revo-
lution, which began in the 1960s and is
sometimes described as the ‘green revo-
lution’, was based on improved plant-
breeding techniques and the development
of hybrid varieties; it now includes genetic
engineering of plants but also a heavy
reliance on the use of chemicals. We may
now be at the cusp of a third stage, which
will combine both approaches in a more
holistic and elegant strategy. Applying
knowledge about beneficial plant–
microbe interactions in the rhizosphere to
plant-breeding and genetic-engineering
technologies may allow us to increase food
production while reducing stress on the
environment and on global biodiversity.
ACKNOWLEDGEMENTS
Research in the authors’ laboratories is supported
by research grants from Science Foundation
Ireland (02/IN.1/B1261; 03/IN3/B373), the HEA
PRTLI programmes, Enterprise Ireland
(SC/02/0420; SC/02/517), the European Union
(QLK3-2001-00101; QLK5-CT-2002-00914) and
the Health Research Board (RP76/2001).
EMBO reports VOL 5 | NO 10 | 2004 9 2 5
 science & society
REFERENCES
Arnold AE, Mejia LC, Kyllo D, Rojas EI, Maynard
Z, Robbins N, Herre EA (2003) Fungal
endophytes limit pathogen damage in a
tropical tree. Proc Natl Acad Sci USA 100:
15649–15654
Bloemberg GV, Lugtenberg BJ (2001) Molecular
basis of plant growth promotion and
biocontrol by rhizobacteria. Curr Opin Plant
Biol 4: 343–350
Bouarab K, Melton R, Peart J, Baulcombe D,
Osbourn A (2002) A saponin-detoxifying
enzyme mediates suppression of plant
defences. Nature 418: 889–892
Clay K (2004) Fungi and the food of the gods.
Nature 427: 401–402
de Weert S, Vermeiren H, Mulders IH, Kuiper I,
Hendrickx N, Bloemberg GV, Vanderleyden J,
De Mot R, Lugtenberg BJ (2002) Flagella-
driven chemotaxis towards exudate
components is an important trait for tomato
root colonization by Pseudomonas
fluorescens. Mol Plant Microbe Interact
15:1173–1180
Friedman M (2002) Tomato glycoalkaloids: role
in the plant and in the diet. J Agric Food Chem
50: 5751–5780
Harman GE, Howell CR, Viterbo A, Chet I, Lorito
M (2004) Trichoderma species—
opportunistic, avirulent plant symbionts. Nat
Rev Microbiol 2: 43–56
Helgason T, Daniell TJ, Husband R, Fitter AH,
Young JPW (1998) Ploughing up the wood-
wide web? Nature 394: 431
Hiltner L (1904) Uber neue Erfahrungen und
Probleme auf dem Gebiete der
Bodenbakteriologie. Arbeiten der Deutschen
Landwirtschaftgesellschaft 98: 59–78
Korpan YI, Nazarenko EA, Skryshevskaya IV,
Martelet C, Jaffrezic-Renault N, El’skaya AV
(2004) Potato glycoalkaloids: true safety or
false sense of security? Trends Biotechnol 22:
147–151
9 2 6 EMBO reports VOL 5 | NO 10 | 2004
v iew point
Lodwig EM, Hosie AH, Bourdes A, Findlay K, Walsh UF, Morrissey JP, O’Gara F (2001)
Allaway D, Karunakaran R, Downie JA, Poole Pseudomonas for biocontrol of
PS (2003) Amino-acid cycling drives nitrogen phytopathogens: from functional genomics to
fixation in the legume–Rhizobium symbiosis. commercial exploitation. Curr Opin
Nature 422: 722–726 Biotechnol 12: 289–295
Loh J, Pierson EA, Pierson LS III, Stacey G, Wardle DA, Bardgett RD, Klironomos JN, Setala
Chatterjee A (2002) Quorum sensing in plant- H, van der Putten WH, Wall DH (2004)
associated bacteria. Curr Opin Plant Biol 5: Ecological linkages between above ground
285–290 and below ground biota. Science 304:
Marcel G, Van Der Heijden A, Klironomos JN, 1629–1633
Ursic M, Moutoglis P, Streitwolf-Engel R,
Boller R, Wiemken A, Sanders IR (1998)
Mycorrhizal fungal diversity determines plant
biodiversity, ecosystem variability and
productivity. Nature 396: 69–72
Morrissey JP, Osbourn AE (1999) Fungal
resistance to plant antibiotics as a mechanism
of pathogenesis. Microbiol Mol Biol Rev 63:
708–724
Newton JA, Fray RG (2004) Integration of
environmental and host-derived signals
with quorum sensing during plant–
microbe interactions. Cell Microbiol 6:
213–224
Nosengo N (2003) Fertilized to death. Nature
425: 894–895
Puhler A, Arlat M, Becker A, Gottfert M,
Morrissey JP, O’Gara F (2004) What can
bacterial genome research teach us about
bacteria–plant interactions? Curr Opin Plant
Biol 7: 137–147
Raaijmakers JM, Vlami M, de Souza JT (2002)
Antibiotic production by bacterial biocontrol
agents. Antonie Van Leeuwenhoek 81: John P. Morrissey (top left) is at the Department
537–547 of Microbiology, University College Cork, Ireland.
Reay DS (2004) Fertiliser ‘solution’ could turn
J. Maxwell Dow (top right), G. Louise Mark and
local problem global. Nature 427: 485
Fergal O’Gara are at the BIOMERIT Research
Smith SE, Read DJ (1997) Mycorrhizal Symbiosis
2nd edn. San Diego, CA, USA: Academic Centre in the Department of Microbiology at
van Loon LC, Bakker PA, Pieterse CM (1998) University College Cork, Ireland.
Systemic resistance induced by rhizosphere E-mail: f.ogara@ucc.ie
bacteria. Annu Rev Phytopathol 36:
453–483 doi:10.1038/sj.embor.7400263
©2004 EUROPEAN MOLECULAR BIOLOGY ORGANIZATION